INTRODUCTION

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

Volvox carteri is a spherical multicellular alga with many features that recommend it as a model for studying the process of cytodifferentiation (Kirk and Harper, 1986) and the early development of photoreception in eucaryotes. Individuals of this species contain only two distinct cell types, 16 large reproductive cells (gonidia) and 2000-4000 somatic cells that cannot divide. The somatic cells are arranged in a single layer at the surface of the transparent sphere, whereas the 16 gonidia are located below the surface, where they have no direct contact with the external medium (Fig. 1; Kirk et al., 1991). All somatic cells are flagellated and possess eyes, and they are responsible for guiding the colony to places of light conditions that are optimal for photosynthetic growth. The orientation of the individual somatic cells within the spheroid, combined with the three-dimensional pattern in which their flagella beat, cause the spheroid to rotate in a counterclockwise direction (as seen along the swimming vector; Hoops, 1993). The two flagella of each cell beat synchronously and in an almost precisely parallel fashion. The flagella of all cells beat toward the posterior of the spheroid and slightly to the right, causing the spheroid to rotate to the left as it moves foreward (Hoops, 1993, 1997). Anterior cells possess larger and more sensitive eyes than posterior ones (Sakaguchi and Iwasa, 1979; Hoops, 1997). In Volvox the photophobic response involves a cessation of flagellar movement, and not a switch to a different beating mode of the sort that causes slow backward movement in most of its unicellular relatives. When a Volvox spheroid is illuminated from one side, its rotation causes the cells to pass repeatedly between the shaded and the lit sides, with the consequence that their flagella slow down or accelerate beating, turning the colony either toward or away from the light (Foster and Smyth, 1980). Whether cells accelerate or decelerate in response to on and off stimuli depends on the light intensity and its illumination history. Thus, in other words, colonial algae orient in light by a complex differential response of the cells at different sides of the colony and not by a differential response of the two flagella in an individual cell. Because algal colonies rotate more slowly than single-cell species, light-mediated signaling in an alga that exists in colonies is also expected to be slower than signaling in a single-celled alga.

View larger version (62K): [in this window] [in a new window]   FIGURE 1   The two cell types in Volvox carteri. Schematized arrangement of somatic cells (S) and gonidia or the embryos derived from the gonidia (G/E) within a Volvox wild-type colony. Somatic cells and gonidia are embedded in the cellular zone (CZ), whereas the deep zone (DZ) and boundary zone (BZ) are free of cells (Kirk et al., 1986).

All light-induced behavioral responses show action spectra with maxima between 490 and 520 nm (Schletz, 1976; Sakaguchi and Iwasa, 1979), suggesting that in Volvox a rhodopsin similar to the chlamyrhodopsin of its unicellular relative Chlamydomonas reinhardtii (Deininger et al., 1995) serves as the functional photoreceptor for behavioral light responses. A volvoxopsin gene (vop) with striking homology to the chlamyopsin sequence (cop) was recently identified (Ebnet et al., manuscript submitted for publication).

We present the first measurements of photoreceptor currents produced by somatic cells of the colony-forming alga V. carteri. Because all cells of the wild-type colony are enclosed in a complex extracellular matrix that precludes this kind of electrophysiological measurement, the studies were performed with a "dissolver" mutant of V. carteri that develops as a suspension of single cells that are substantially free of such a matrix. Photocurrents were recorded from intact cells and excised eyes in response to short flashes or longer light pulses under various ionic conditions. All recorded photocurrents are well explained by two independent conductances that are localized within the eyespot area.

	MATERIALS AND METHODS

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

Description of the Volvox mutant used for photocurrent recording

In a wild-type V. carteri spheroid, all cells are enclosed in a complex extracellular matrix that surrounds all cells and holds them in fixed relationship to one another (Fig. 1). Therefore our electrical studies were performed with a "dissolver" mutant that develops as a suspension of single cells free of extracellular matrix (ECM). The dissolver strain used here (W251) (gls/regA/dis) arose as a spontaneous mutant in a culture of a previously described "gonidia-less/regenerator" strain (Tam and Kirk, 1991) and was provided to us by Dr. D. Kirk (St. Louis, MO). In wild-type V. carteri there is a complete division of labor in which the biflagellate somatic cells are specialized for motility and are incapable of dividing, whereas the gonidia are specialized for reproduction and never have functional flagella. In the Gls/Reg strain, however, this division of labor is abolished by two mutations (gls and regA): all cells of a Gls/Reg mutant first develop as biflagellate cells that are indistinguishable from wild-type somatic cells in their phototactic activity; then a day later they all dedifferentiate (resorbing their flagella and eyespots) and redifferentiate as gonidia that will divide to produce progeny of like phenotype. With the addition of a third mutation (dis) that interferes with the production of the extracellular matrix (ECM), which normally functions to hold the cells together after they completed embryogenesis, the W251 (Dis/Gls/Reg) strain develops as a uniform population of cells that separate from one another at the end of embryogenesis. These individuals first develop as somatic cells with flagella and eyespots and later differentiate as eye- and flagella-less gonidia. The W251 mutant was grown synchronized under a 16h/8h light-dark regime (14 Wm²) in standard medium (Provasoli and Pinter, 1959) at 28°C.

Electrical measurements

For electrical measurements cells of the mutant strain W251 were centrifuged at 3000 rpm and resuspended in NMG⁺/K⁺ buffer (5 mM HEPES, 10 mM Cl, 1 mM K⁺, 200 µM 1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid, 300 µM Ca²⁺, adjusted with N-methyl-D-glucamine (NMG) to pH 6.8). After dark adaptation for more than 1 h, cells were used for experiments. The NMG⁺/K⁺ buffer was used as the electrode and bath solution. Ca²⁺ was added as CaCl₂. The total Ca²⁺ concentration required for a defined concentration of free Ca²⁺ was calculated according to the method of Holland et al. (1996).

Suction pipette measurements

Photocurrents were recorded by using borosilicate suction pipettes with a final tip diameter in the range of one-half of the cell diameter (i.d.) (Harz et al., 1992; Holland et al., 1996). The pipettes had an access resistance of 20-50 M. Cells were sucked into the pipette by up to one-half until the resistance reached 60-150 M. Under these experimental conditions, one-third of the total current can be detected under the capacitive mode (for further explanation see Holland et al., 1996). Currents were recorded at constant voltage (0 mV between bath and pipette) and were filtered with a 3-kHz low-pass Bessel filter. Data were recorded and processed as described by Harz et al. (1992). If not otherwise indicated, the current traces shown are the mean of 7-10 individual recordings filtered with a digital Gaussian filter to 500 Hz. The orientation of the cells was not optimized for maximum light sensitivity as described before by Harz et al. (1992).

Patch pipette measurements

Pipettes for measuring P-currents directly at the eyespot were pulled from borosilicate glass capillaries (1.8-mm outer diameter, 0.15-mm walls, Kimax-51; Witz Scientific, Maumee, OH) in two steps and were polished until the tip diameter reached ~1.5 µm. The cone angle was ~30°. The pipettes were filled and currents were measured in NMG⁺/K⁺ buffer. The resistance of the pipette was 15-20 M. When the eye (diameter 1.5 µm) was sucked into the pipette, the resistance increased to 120-160 M. A 40× objective (NA = 1.3; Achrostigmat, Zeiss) and a 4× phototube were used for identifying the eyespot in infrared light on the screen. Starting from this configuration, eyespot vesicles were prepared. While low pressure was applied to the eye in the pipette, the major part of the cell was cut off with a second pipette (see Fig. 5). The remaining eyespot vesicle was pressed against the pipette tip by applying weak pressure to create an acceptable seal resistance (120 M). The eyeless cell was kept in the second pipette for control experiments. Currents were filtered with a 10-kHz low-pass Bessel filter and recorded with a frequency of 100 kHz.

The cells or vesicles were stimulated through the objective by a 10-µs flash (IG&G FXQG-949-1; Polytech, Waldbronn). One hundred percent photon exposure (500 nm, 60 nm half-bandwidth) corresponds to 2.6 × 10²⁰ photons m² in the objective plane. Light pulses of 500 ms were applied with a 75-W xenon lamp, and the duration was controlled with a fast electronic shutter (Uniblitz model T132; Vincent Associates, Rochester, NY). One hundred percent photon irradiance (500 nm, 60 nm half-b.w.) corresponds to 1 × 10²¹ photons m² s¹.

If not otherwise indicated, the measurements were carried out at room temperature (20°C).

	RESULTS

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

The Dis/Gls/Reg strain, used in all electrical measurements described below, provided two advantages. First, it produced biflagellate somatic cells that could easily be drawn into a suction pipette. The cell surfaces are sufficiently free of extracellular matrix (ECM) that photoreceptor currents were readily examined and characterized. Second, the ability of these somatic cells to redifferentiate as eyeless gonidia provided us an opportunity to determine whether the gonidia continue to produce photocurrents after they have lost their visible eyespots, which are known to serve as an optical system (Foster and Smyth, 1980).

Recording photocurrents from the Dis/Gls/Reg mutant

To track down rhodopsin-mediated electrical events, small somatic cells with a diameter of 8 µm were sucked into the suction pipette with flagella and eyes outside the pipette and were stimulated with bright green flashes (>10²⁰ photons m²). Transient photocurrents with a peak amplitude of up to 12 pA appeared (fast photoreceptor current, P_F, in Fig. 2 a). The photocurrents peaked at ~2.5 ms after the flash and decayed with  in the range of 12 ms to a new plateau. The current finally decayed within 0.5-1 s (slow photoreceptor current, P_S). Photocurrents recorded from larger Volvox cells with diameters of 10-16 µm had a very similar peak amplitude (Fig. 2 b). Only the decay was slightly retarded. The  values and the integrals increased up to 1.6-fold. When the cells reached diameters above 20 µm, the eyespot pigmentation disappeared and the cells turned into gonidia and began to divide. Independent of the number of divisions completed, neither the gonidia nor their daughter cells exhibited any photocurrents (Fig. 2 c). Apparently, the electrical photoresponsiveness became totally inactivated during conversion into gonidia within a time range of 2-3 h. Such an abrupt, developmentally controlled inactivation of a visual process has not been observed, to our knowledge, in any other organism.

View larger version (67K): [in this window] [in a new window]   FIGURE 2   Cell specificity of the photocurrents. Normarski images (DIC), chlorophyll fluorescence, and flash-induced photocurrents (PF and PS) were recorded from strain W251 small and large somatic cells (a and b) and from a reproductive cell (gonidium) after first division (c). Currents were recorded from somatic cells with eyespot and flagella outside the pipette. Bar = 10 µm.

Light dependence of the P-currents and evidence for its activation by rhodopsin

The photocurrents recorded from the somatic cells are all confined to the eyespot region. The photocurrents had a positive sign under our measuring conditions whenever the eyespot was outside the pipette, and there was an inward current from the bath into the eyespot (Fig. 3 a). When the eyespot was in the pipette, the inward current from the pipette into the eye was monitered, and the whole current inverted to negative (Fig. 3 b; Harz et al., 1992; Holland et al., 1996), indicating that all current components are localized within the eyespot area. It is expected that the P-currents are larger when measured with the eyespot in the pipette compared to recordings with the eyespot outside, because a large fraction of the total current is detected. However, this was not the case, because the eyespot was in most experiments pressed against the pipette wall and was not freely accessible. The kinetic features of the currents were identical in the two eyespot orientations. The kinetics of the maximum response, shown in Fig. 3 b, were determined as ₁ = 1.0 ms, ₂ = 12.7 ms, and ₃ = 76.4 ms. P_F was graded with the photon exposure (Fig. 3, a and b). To study photocurrents at low photon exposure, we used patch pipettes with a small tip diameter and a steep cone angle (Fig. 3 c). This configuration improved the resolution and allowed us to record photocurrents of less than 0.2 pA. The stimulus-response curve of the P_F peak covered a dynamic range of almost four log units of photon exposure (Fig. 3 c). The slope was too small to be fit by an exponential or by a hyperbolic saturation curve.

View larger version (24K): [in this window] [in a new window]   FIGURE 3   Light dependence. (a and b) Flash-induced fast (PF) and slow (PS) photoreceptor currents recorded with a suction pipette from somatic cells with eyespots outside (a) or inside (b) the pipette. A 100% flash corresponds to a photon exposure of 2.6 × 1020 photons m2. The solid line in b represents a fit. The rise is monoexponential (1 = 1.0 ms), whereas the decay is biexponential (2 = 12.7 ms and 3 = 76.4 ms). (c) Stimulus-response curves of normalized PF peak amplitudes () and the linear phase of the current rise (). Each data point represents the average of seven recordings. The dashed line shows a saturation curve calculated according to Eq. 1. The solid line is the curve fitting according to Eq. 1. (d) Normalized PF currents recorded at 100%, 50%, and 27% photon exposure from the same cell after substraction of the 3% trace shown under b. (e) Plot of the Volvox PF integral versus the normalized PF amplitude () in comparison with integrals of the P current recorded from a Chlamydomonas cell (strain CW2) of the same size (). (f) Action spectrum of the PF current recorded with patch pipettes directly from the eye () in comparison with a published low-intensity action spectrum for photophobic responses () (data taken from Schletz, 1976) and a high-intensity spectrum of the same response () (data from Sakaguchi and Iwasa, 1979) of wild-type Volvox carteri colonies. The data points of the PF spectrum are averaged values of four cells.

To study P_F independently of the P_S, the trace recorded at 3% photon exposure (Fig. 3 b), where only a small fraction of the rhodopsin is bleached, was substracted from the traces recorded at higher flash energies. The residual currents were normalized; these are shown in Fig. 3 d. The rise, the peak time, and the decay of these residuals were almost constant for every cell and independent of the current size. The decay was now described by a single exponential, and the -value was close to ₂ of the total current shown in Fig. 3 b. Because of the increasing amplitude and the constant decay, the current integral is graded with the photon exposure (Fig. 3 e), and because the cell is almost at isopotential, the degree of depolarization should change accordingly (Harz et al., 1992). The depolarization must be smaller in larger cells because the P-current integral does not increase in parallel with the cell surface area. The invarability of the P_F time constants is a clear difference from Chlamydomonas, where P-currents with large amplitude decay faster than smaller currents, leaving the P-current integral and the degree of depolarization rather constant between 10% and 100% receptor bleaching (Fig. 3 e).

The linear part of the P_F stimulus-response curve was determined at five different wavelengths. After the curves were extrapolated to zero, the threshold sensitivities were plotted as a function of photon energy (Fig. 3 f). The action spectrum peaked at 495 nm (2.51 eV). The close fit of the spectrum to a rhodopsin standard curve on one hand and to the action spectrum for rhodopsin-triggered photocurrents in Chlamydomonas on the other hand leaves little doubt that the photoreceptor is a rhodopsin. The spectrum matches, as expected, the low-intensity action spectrum for phototaxis, whereas the high-intensity phototaxis spectrum is finely structured (Fig. 3 f). The shape of the photocurrent action spectrum and its location on the energy scale exclude the contribution of phytochrome as well as flavin-based light receptors (Smyth et al., 1988).

The cells did not exhibit any action potential like flagellar currents, F_F, which is a great benefit for the analysis of the photoreceptor currents. In single-celled algae F_F overlaps with the P-current and distorts the measured decay at high flash energies (Litvin et al., 1978; Harz and Hegemann, 1991).

The light dependence of the P-current rise

In flash experiments the P-current rise was determined by extrapolating its linear part to zero (Fig. 4 a, inset). The rise increases with the photon exposure, and its stimulus-response curve is fit by a single exponential function quite adequately (Fig. 3 c). An exponential light dependence is expected if the fraction of channels activated and the number of charges entering the cell per time unit are proportional to the photon exposure until all rhodopsin is bleached. The simplest explanation is a constant ratio between actived receptor molecules and the number of activated channels:

(1)

where P is the number of channels activated, and Q is the amount of light absorbed (photon exposure) (Lamb et al., 1981). A hyperbolic saturation curve, Qk/(1 + Qk), describes the experimental data less accurately, even if a Hill coefficient is included (data not shown). Substituting in Eq. 1 the constant k for the product of a typical rhodopsin absorption cross section  = 1.9 × 10²⁰ m² times the rhodopsin quantum efficiency of  = 0.67, numerical values were calculated. This calculated saturation curve is shifted to lower flash energies by only a factor of 2 relative to the experimental data.

View larger version (21K): [in this window] [in a new window]   FIGURE 4   Delay and flash-to-peak time. (a) The delay between flash and beginning of the Volvox P current is plotted versus log photon exposure (). The delay of the P current of the unicellular alga Chlamydomonas was measured under identical conditions and plotted for comparison () (averages of 50 individual recordings, 100-kHz sampling, 10-kHz low-pass filter, 10-kHz Gauss filter). (b) The flash to peak time is plotted versus log photon exposure (averages of 10 individual recordings, 10-kHz low-pass filter, 1-kHz Gauss filter). All traces were measured with an eyespot inside a small patch pipette. The insets show representative responses to flashes of the indicated energies. The arrows in the insets mark the time of the flash.

In Volvox, the P_F induced by an intense flash of 10 µs duration and a photon exposure between 10¹⁹ and 10²⁰ photons m² begins to rise with virtually no delay (below 50 µs; Fig. 4 a), which closely resembles the rapid P-current rise in Haematococcus and Chlamydomonas (Sineshchekov et al., 1990; Holland et al., 1996). This immediate channel activation led to the suggestion that the rhodopsin and the photoreceptor channel in microalgae are in close contact or even form one functional complex (Harz et al., 1992). Now, because of the improved signal resolution, the Volvox P-currents were measured down to flash energies of only 10¹⁶ photons m², where only a small fraction of rhodopsin molecules (0.01%) are activated. At this low photon exposure the current rises in a sigmoidal fashion. The delay, also defined by extrapolating the linear part of the curve to zero (Fig. 4 a, inset), increases dramatically at low photon exposure, reaching in Volvox values up to 10 ms when only a total of few photons are absorbed (Fig. 4 a). The flash-to-peak time is extended accordingly from 2.5 and 25 ms (Fig. 4 b). In contrast to the delay, the variability of the flash-to-peak time is detectable up to light saturation. It is well explained solely by the light dependence of the rise shown in Fig. 3 c. Both the P_F delay and the flash-to-peak time are larger than their counterparts in Chlamydomonas, which are plotted for comparison in Fig. 4.

Photocurrents recorded from excised eyes

When the Volvox cells are sucked into small patch pipettes and strong underpressure is applied, vesicles are sometimes pulled off into the pipette. Better control of the vesicle size is achieved by pinching them off mechanically with the help of a suction pipette (Fig. 5 a). Photocurrents could be recorded from those amputated cells that still contained the eye. Proportionally to the size reduction of the cell, both P-currents were reduced in their amplitude (Fig. 5 b). However, the kinetics of the P-current, the time of the current maximum (Fig. 5 c), and the light sensitivity of the cell (not shown) were left unchanged. Finally, photocurrents were recorded from vesicles that contained only the eyespot and very little of the residual cell (Fig. 5 b). These small vesicles had to be pressed against the inside of the pipette outlet to produce the same electrical resistance as with intact cells. The amplitude of the P-currents was in the range of only 10% of the currents recorded from intact cells, but all other properties were again left unchanged. The reason for the size reduction is unclear, but certainly the seal between pipette and vesicle contributes to a larger extent to the total resistance, and it is conceivable that the membrane potential of the vesicle is less negative than that of the intact cell. Cells with no eyes were totally insensitive to light stimulation. These experiments provide the most direct proof so far that the receptor and the signal transduction system Volvocales are located within the eyespot area.

View larger version (13K): [in this window] [in a new window]   FIGURE 5   Photocurrents from eyespot vesicles. (a) Preparation. After shearing of the major part of the cell with a suction pipette, the eyeless residual cell was kept in the suction pipette, whereas the eye vesicle remained in the patch pipette. (b) Photocurrents recorded from intact cells (1), from cells with reduced volume (2), and from a small eye-containing vesicle (3) (photon exposure: 2.6 × 1020 photons m2). Eyeless cells were unresponsive to light. (c) Normalized currents taken from b, 1 and 2.

The slow photoreceptor current, P_S

As seen in Figs. 2, 3 and 5, the P_F is followed by a second current component that decays only slowly. The light dependence of this P_S is different from that of the P_F current. Its amplitude under physiological conditions (200 µM Ca²⁺, 1 mM K⁺, pH 7) is only 0.5-1 pA, with little change over a range of two orders of photon exposure (Fig. 6 a). In other words, it is already saturated when only a few percent of the photoreceptor is bleached. It should be kept in mind that this relationship is independent of the number of photoreceptors per cell. In contrast, the duration of P_S continuously increases with the photon exposure in a dose-dependent manner (Fig. 6 b), i.e., when more receptor is bleached the duration but not its size increases.

View larger version (12K): [in this window] [in a new window]   FIGURE 6   The slow photoreceptor current PS. (a) Light dependence of the P currents, with emphasis on slow PS current duration. Arrows indicate when PS falls below 0.5 pA. (b) The duration of PS plotted versus log photon exposure.

The stationary photoreceptor current, P_St

Because the natural stimulus of a Volvox colony is modulated continuous light instead of short flashes, in a separate set of experiments light pulses with a duration in the second range were applied to small somatic cells. Besides the transient P_F current, a stationary current appeared and continued as long as the light was kept on (at least for 20 s) (Fig. 7 a). Modulation of the photon irradiance during the light period modulated the amplitude of the stationary current accordingly (data not shown). The current sign inverted from positive to negative when the eye was sucked into the pipette (as seen in Fig. 7 a), independent of the location of the flagella. Because this sign inversion indicates the eyespot location, the stationary current is again an eye-specific current and was termed P_St. The P_St amplitude saturates at low photon irradiance (Fig. 7 b) in a way similar to that of the peak amplitude of P_S. P_F only appeared when P_St nearly saturated. After the light was turned off, P_St decayed with a time constant  of around 150 ms, independent of the former light intensity.

View larger version (16K): [in this window] [in a new window]   FIGURE 7   Stationary photocurrents. (a) Photocurrents in response to a light pulse (100% and 25%) with the eyespot inside and flagella outside the pipette. The duration of the light pulse is indicated above the current traces (12 averages, digital Gaussian filter 100 Hz). (b) The amplitudes of PF and PSt are plotted versus log photon irradiance (mean values from five cells).

Ca²⁺, H⁺, and K⁺ dependence of the P currents

The photocurrents depend in a distinctive way on the extracellular ion composition. Lowering the extracellular Ca²⁺ below 10 nM in flash experiments led to a disappearence of the transient P_F within 1 min, whereas P_S remained unchanged. P_F immediately reappeared when Ca²⁺ was readded and the flashing was continued (Fig. 8 a). Apparently, under physiological conditions P_F is mainly carried by Ca²⁺, whereas the P_S is Ca²⁺-independent. Ba²⁺ substitutes for Ca²⁺ ions with unchanged kinetics, suggesting that Ca²⁺-regulated intracellular processes are not involved. At 200 µM Ca²⁺, the addition of 1 mM La³⁺ led to a total disappearence of all electrical light responses. The stationary current, P_St, is not dominated by Ca²⁺ but shows some Ca²⁺ dependence. Its size varied by less than a factor of 2 when the extracellular Ca²⁺ concentration was varied between 10 nM and 200 µM (Fig. 8 b). Thus a Ca²⁺ conductance might contribute to P_St as in animal photoreceptor currents, but it is certainly not the major current carrier.

View larger version (18K): [in this window] [in a new window]   FIGURE 8   Ca2+ and pH dependence. (a and b) Suppression of the PF current upon reduction of Ca2+ from 200 µM (+), to below 10 nM () in the bath solution, in a flash experiment (a) and in an experiment where light pulses were applied (b). The duration of the light pulse is indicated above the electrical traces (K+ = 1 mM, pH 6.8). (c and d) Photocurrents recorded at different pH values at high Ca2+ (200 µM (c)) and low Ca2+ (< 10 nM (d)). Note that the time scale in a is different from that in b-d. Currents were filtered to 100 Hz (b and c) or 50 Hz (d) with a digital Gaussian filter.

All three currents, P_F, P_S, and P_St, are profoundly affected by the extracellular pH. At acidic pH both P_F and P_S dramatically increase in size, suggesting that both currents are mainly carried by H⁺ under these conditions (Fig. 8 c). The kinetics of P_F remained stable, whereas the duration of P_S was largely extended. To exclude the possibility that the currents are acid-facilitated Ca²⁺ influxes, the experiments were repeated in the absence of Ca²⁺ (Fig. 8 d). When the pH was changed from 6.8 to 4.1, the same dramatic increase in P_S and P_F reappeared, suggesting that both components are carried by H⁺ under acidic conditions. In continuous light the P_St amplitude was enlarged in a manner similar to that of P_S in flash experiments (data not shown).

P_F, P_S, and P_St differ from each other by their K⁺ sensitivity. P_F is independent of the extracellular K⁺ concentration over a wide concentration range between 100 µM and 10 mM, provided the cells have adapted to the K⁺ change. This observation is consistent with the interpretation that P_F is not accompanied by significant K⁺ efflux. The K⁺ conductance, G_K, of the plasmalemma is low but not negligible (Malhotra and Glass, 1995). P_S currents were unaffected by K⁺ at pH 7 and low flash energies, but the high-energy responses were reduced by K⁺. The enhanced P_S currents recorded at pH 4 were reduced at elevated K⁺ at all flash energies. P_St currents completely disappeared at 10 mM K⁺ (Fig. 9 a). These findings are consistent with a light-induced K⁺ conductance, G_K. The K⁺ efflux promotes P_S and P_St by stabilizing the membrane potential near E_K. The small positive current that continues beyond the duration of the light pulse at 10 mM K⁺ (Fig. 9 a) is explained as a slightly unevenly distributed K⁺ influx. At asymmetrical extracellular K⁺ with 7 mM K⁺ in the bath and 100 µM in the pipette, a transcellular K⁺ flux is created. K⁺ is directed through the activated K⁺ conductance, G_K, from the bath into the cell and from the cell into the pipette (Nonnengässer et al., 1996). This transcellular K⁺ flux is permanent in continuous light, whereas it is seen as a clearly defined peak that follows P_F in flash experiments (Fig. 9 b). In Chlamydomonas G_K is activated after the fast flagellar current (Govorunova et al., 1997), whereas in Volvox G_K seems to be activated by the P_F-induced depolarization (Fig. 9 b).

View larger version (21K): [in this window] [in a new window]   FIGURE 9   K+ dependence. (a) Symmetrical high K+ (10 mM) in bath and pipette abolishes the K+ efflux, rendering PSt undetectable as schematized in the inset (five averages, Gaussian filter 100 Hz). (b) Visualization of the PF-triggered K+ efflux at unsymmetrical extracellular K+. The 7 mM K+ in the bath solution and 100 µM in the pipette direct the K+ into the pipette (average of 23 responses, 100 Hz Gauss).

	DISCUSSION

Top Abstract INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

Rhodopsin activation of the photoreceptor currents

The above-described experiments demonstrate in the spheroidal alga Volvox carteri the existence of photocurrents, which have all the characteristics needed to qualify them as the trigger for behavioral responses. The rhodopsin shape of the P-current action spectrum and its peak at 495 nm leave little doubt that the P-currents are triggered by rhodopsin. As already mentioned, the published action spectra for movement responses in colony-forming Volvocaceae all peak between 490 and 534 nm (Mast, 1917; Luntz, 1931; Halldal, 1958; Schletz, 1976; Sakaguchi and Iwasa, 1979). Reasons for the differences among the observed maxima are the different recording techniques and the intensities at which the responses were measured. In general, low-intensity action spectra for phototaxis, as well as the spectra for flash-induced phobic responses (flagellar arrest), peak at 490-500 nm and are rhodopsin shaped (Luntz, 1931; Schletz, 1976). They are in agreement with the photocurrent action spectrum of Fig. 3 f. The correlation supports the claim that the P-currents are the trigger for phobic responses and for phototaxis at low light. In contrast, high-intensity (finite response) phototaxis spectra are structured and red shifted by 20-30 nm. But these differences are explained by the photoreceptor optics, by shading pigments, by adaptation phenomena, and potentially by photoreceptor photochromism (Foster and Smyth, 1980; Hegemann and Harz, 1998). Thus the correlation between high-intensity action spectra and absorption of the rhodopsin is less accurate.

Classification of the conductances

The major properties of the observed photocurrents are summarized in Table 1. Until now nothing was known about the visual process in multicellular Volvocacean algae such as Volvox, Pandorina, Eudorina, and other relatives. Only the availability of dissolver mutants made it possible to obtain information about the electrical processes. The Volvox spheroid rotates at 0.45 Hz (Gerisch, 1959; Schletz, 1976; Sakaguchi and Iwasa, 1979), which is slow compared to the 2 Hz rotation frequency of most unicellular flagellates (at room temperature). This explains why in Volvox all photocurrents are slower than in single-celled flagellates and why there is no need for a faster signaling system. One should keep in mind that a Volvox cell in nature always swims and navigates as part of a large spheroid and never as a single individual. The relatively slow P-current rise makes this alga a preferable model system for studying the rhodopsin-ion channel coupling in flagellate algae.

The shape of the photoreceptor currents could be explained in at least three ways. First, the P_S plateau is attributable to a desensitisation of the transduction mechanism; or, second, the current is limited by an outwardly directed countercurrent across the plasmalemma (K⁺ efflux); or, finally, the transient P_F and the slowly decaying P_S reflect two independent conductances.

All Volvox photoreceptor currents (P-currents) are localized inward currents into the eyespot region, whereas the repolarizing K⁺ efflux is more or less evenly distributed over the plasmalemma. The ion that carries the P_F has not been unequivocally identified. But the sensitivity to extracellular Ca²⁺, the maintenance of the conductance by Ba²⁺, and its inhibition by Ca²⁺ channel inhibitors are cumulative evidence that P_F is mainly carried by Ca²⁺. Neither the rapid response to external Ca²⁺ nor the Ba²⁺ substitution is compatible with the ideas that Ca²⁺ only activates the channel from inside via Ca²⁺ stores (Plieth et al., 1998) and that the channel conducts other ions but not Ca²⁺. But the strong increase in the P_F current at acidic pH without any change in the kinetics suggests that the underlaying conductance, G_Ca, conducts H⁺ as well under selected conditions. In contrast, P_S is quite obviously carried solely by H⁺. A direct proportionality of the current amplitude to the extracellular H⁺ concentration over a wide pH range is not expected, because the cell depolarizes during the H⁺ influx and the K⁺ efflux should become rate limiting at high H⁺. Nevertheless, P_F and P_S must be mediated by two independent conductances, G_Ca and G_H, similar to the two conductances in Drosophila eyes (Niemeyer et al., 1996).

Stationary photocurrents like P_St were observed before in the unicellular flagellate Haematococcus (Sineshchekov, 1991), but the amplitude was small and the nature of this current is unknown. The authors stated that the ion of the current is clearly not calcium, because it is not inhibited by Ca²⁺ removal. In Chlamydomonas the stationary current is of similarly small size (Braun, unpublished observations). In Volvox, stationary P-currents are large, producing a massive cation influx during extended light pulses. The above experiments are consistent with a contribution of both conductaces G_Ca and G_H to P_St, at which G_H is dominating under most conditions. The influx may reach 1 × 10⁷ charges/s (at pH 7), corresponding to a total concentration increase of 10 µM H⁺/s in a 15-µm cell.

A contribution of anions as Cl to any of the photocurrents is electrically possible because the equilibrium potential for Cl should always be positive, but so far there is no single experiment at hand that suggests an anion efflux.

Channel activation

There are two photon exposure ranges that can be discriminated with respect to photocurrent activation. At the high intensity range above 10¹⁹ photons m², the delay is below 50 µs. The extremely rapid activation of the P-current after a light flash in flagellate algae, first observed in Haematococcus (Sineshchekov et al., 1990), led to the suggestion of a direct coupling between rhodopsin and the ion channel (Harz et al., 1992). This close link was now confirmed in Volvox. However, the hypothesis remained controversial because in all former studies the peak amplitude was analyzed and the light dependence did not follow a monoexponential function (Sineshchekov, 1991; Harz et al., 1992; Beck, 1996). The use of pipettes with steep cone angle now improved the resolution and made it possible to study on one hand photocurrents in response to very low flash energies, and on the other hand to analyze the P-current initial rise. The saturation of the rise follows a single exponent, as expected for a fixed coupling between rhodopsin and the channel. A linear signal chain explains the sigmoidicity of the rise if the light-excited rhodopsin activates the channel via at least one intermediate state: This intermediate state, I, may be an intermediate of the rhodopsin photocycle, a separate transmitter compound, a separate state of the channel, or, finally, a state of a common rhodopsin-ion channel complex. However, in contrast to a signaling system with dynamic amplification, a linear model with fixed decay rates cannot explain the light dependence of the delay between flash and P-current rise (Gradmann and Hegemann, unpublished observations).

The flat slope of the stimulus-response curve for the peak current was explained for Chlamydomonas by the depolarization during the response. This depolarization accelerates the current decay, shifts the peak to shorter times, and reduces the current size at high flash energies (Harz et al., 1992). In vertebrate vision, flash-induced currents are quite independent of the voltage changes occurring during a flash response (Bader et al., 1979), but nevertheless the influence of voltage changes is visible (Lamb et al., 1981). However, depolarization does not explain the flat dose-response curve for peak currents in Volvox. Here the P_F decay is independent of the current size, which would not be the case if the peak current were reduced by the progressing depolarization. Second, because of the slow decay, at peak the current integral reached only 10% of the total. In a Volvox cell 15 µm in diameter, P-currents with integrals of up to 300 fC should lead to a depolarization of ~50 mV (for a detailed explanation see Harz et al., 1992). Thus the depolarization is only 5 mV at peak and cannot explain a strong size reduction at high flash energies due to voltage inactivation.

The exponential saturation curve of the P-current rise and its close fit by the numerical calculation based on typical rhodopsin values implicates that the P_F rise is directly proportional to the amount of rhodopsin bleached by the flash. It is also compatible with the earlier presented model that the rhodopsin and the primary channel form a single protein complex. However, the nonexponential saturation curve for the total peak current, the biphasic current decay with a light dependent ratio of ₁/₂, and the different ion dependence of P_F and P_S are, taken together, only explained by two signaling systems as they are schematized in Fig. 10. A saturating flash of 4-8 × 10²⁰ photons m² bleaches practically all rhodopsin molecules of the cell. The bleached photoreceptors undergo a transition to the signaling state, from which they activate by direct contact a cation conductance, G_Ca, in the eyespot overlaying part of the plasmalemma. It conducts mostly Ca²⁺ under physiological conditions. The activation occurs in the submillisecond range and is faster than ion channel activation in any animal visual system. The Ca²⁺ influx reaches its maximum after 2.5-3 ms. Under all conditions the P_F current of a particular cell decays with the same time constant. Hence the decay rate is an intrinsic property of the channel. Second, rhodopsin induces a long-lasting inward current of only 1 pA. This P_S current is carried by H⁺. P_S is activated with a delay of up to 10 ms and peaks after 2.5-25 ms. The responsible channel is not directly activated by rhodopsin but rather via a signal amplification system, which is likely to include a G-protein. The conductance G_H saturates when only a few percent of the rhodopsin is bleached.

View larger version (46K): [in this window] [in a new window]   FIGURE 10   Model of rhodopsin-triggered conductances. Upon flash stimulation light-excited rhodopsin, Rh*, activates two conductances, GCa and GH, GCa by direct contact and GH most likely via a transducer (T), which could be a G-protein. The corresponding channels PCCa and PCH are localized within the eyespot area. PCCa inactivates spontaneously and independently of the extracellular Ca2+ concentration and the actual membrane potential. The inactivation of PCH is unclear, but most likely is brought about by the decay of the messenger. In continuous light, both GCa and GH in combination with a nonlocalized K conductance, GK, constitute the stationary current PSt.

Because of the chlamyopsin sequence, which predicts a highly charged protein with some sequence homology to K⁺ channels, we had suggested that chlamyrhodopsin is part of a rhodopsin-ion channel complex or even forms a channel by itself, or, in other words, it constitutes a channel with an intrinsic light sensor. The volvoxopsin is up to 62% identical to chlamyopsin and certainly belongs to the same gene family. It exhibits very similar features and should fulfill the same role. On the other hand, if only one type of rhodopsin exists in green algae as concluded by biochemical experiments (Kröger and Hegemann, 1994), the rhodopsins must also function as a remote sensor activating P_S via an amplification system. Despite these needs, there are no evident G-protein-binding domains in the algal opsin sequences (Ebnet et al., manuscript submitted for publication). However, Kreimer and colleagues identified a G-protein in eyespot fractions of the green alga S. similis. Its GTPase acitivity is light dependent and is sensitive to anti-chlamyopsin antibodies (Calenberg et al., 1998). Thus this G-protein is an appropriate candidate for enabling rhodopsin-triggered remote sensing in green algae. G-protein activation and the downstream ion channel modulation occur in animal vision within the expected time scale of 2.5-25 ms (Felber et al., 1996), which is the time scale observed for the P_S delay.

Physiological implications

It is now abundantly clear that in Chlamydomonas as in other eucaryotes from ciliates to mammals, flagellar activity is regulated by the Ca²⁺ level in the cytoplasm (Tamm, 1994). It is also accepted that a Ca²⁺ ATPase operates continuously to keep the Ca²⁺ concentration at a submicromolar level. But from the presented data it is clear now that in green algae during the continuous photocurrents, Ca²⁺ only accompanies monovalent ions as in animal visual processes, but may fulfill its physiological role as the key control element for flagellar beating (Witman, 1994; Holland et al., 1997) without any restriction.

Ca²⁺ is a favorable ion for carrying the P_F because in dark-adapted Volvox cells Ca²⁺ is under a higher driving force. The intracellular Ca²⁺ is in the range of 100 nM (M. Fischer, unpublished) and between 0.1 and 0.3 mM in the growth medium and in the bath during experiments. On the other hand, high cytoplasmic Ca²⁺ levels would be harmful. Thus it is more beneficial for the alga if Ca²⁺ plays a modulating role in continuous light, and H⁺ instead of Ca²⁺ is used as the major current carrier. However, the H⁺ gradient is relatively small at pH 7 (below 1 log unit at pH 7; Malhotra and Glass, 1995; Fischer and Braun, unpublished), and the amplitude is accordingly small. At higher pH gradients the driving force for H⁺ is increased, but the H⁺ influx is only maintained if it is accompanied by the K⁺ efflux.

The finding that the photoreceptor currents increase with the photon exposure until all of the rhodopsin is bleached provides new perspectives for the control of the flagella that execute the response. If the flagellar beat frequency is controlled in Volvox by the membrane potential, by a proton influx, and by Ca²⁺, beating should be modulated over a broad range of photon exposure. Unfortunately, this has not been tested so far. But beating in response to step-up stimulation was measured by video microscopy (reviewed by Hoops, 1997). As expected, the beating was reduced almost to zero and recovered to a stationary level after one or two seconds. The transient frequency reduction is larger compared to that of Chlamydomonas or other single-celled algae. But single-celled species compensate for that deficit with the ability to perform phobic responses (flagellar reprogramming). These shock responses are caused by an action potential-like fast flagellar current, F_F, which causes a massive Ca²⁺ influx along the whole flagellar length, thus triggering backward swimming for some hundred milliseconds (Litvin et al., 1978; Harz and Hegemann 1991; Beck and Uhl, 1994; Holland et al., 1997). This type of response and consequently the F_F current are totally absent in Volvox. Backward swimming makes no sense for an alga that is part of a large colony of several thousand individuals. Ca²⁺ is expected to modulate the flagellar beat frequency by acting only at the flagellar basis (on-off response; Tamm, 1994). This may happen in two ways: either there are Ca²⁺ channels restricted to the flagellar base, which have a conductance too low to be seen in our experiments, or Ca²⁺ diffuses from a cytosolic source to the flagella.